Enhancing Swimming Performance of a Biomimetic Robotic Fish by Optimizing Oscillator Phase Differences of a CPG Model

This paper presents a novel approach on swimming performance improvement for a bio-inspired robotic fish. The main idea of this approach lies in optimizing the oscillator phase differences of a central pattern generator (CPG) model, which governs the rotation of the joints. Definitely, the swimming performance includes steady forward velocity, head stability, and energy-efficiency. A dynamic model is firstly built to describe the swimming motion of the robotic fish. Taking advantage of the dynamic model, we analyze the influence on the swimming performance exerted by the phase differences. Specifically, two conditions are studied and compared together: consistent and inconsistent phase differences. To the best of our knowledge, this issue has not been addressed in previous literature sufficiently. The optimal swimming performance is searched under both of the conditions. Simulation results reveal that the CPG model with inconsistent phase differences is able to realize better swimming performance, especially for head stability and energy-efficiency.